The Oxfordshire Community Stroke Project (OCSP) classification provides a simple means of classifying the clinical syndromes associated with acute stroke. The validity of the classification can depend on many factors. Accuracy and time of the clinical and radiological examination are very important. It was used in the International Stroke Trial (IST). The study was conducted in 467 hospitals in 34 countries. Our aim was to assess how well the OCSP classification could predict infarct site and size on computed tomography (CT) scan when performed in a trial within 48 h after the onset of stroke and the clinical assessment was carried out by different doctors in different hospitals. We examined data on the patients randomized in the IST by the seven participating hospitals in Poland. Patients admitted to the hospital were examined by the doctor on duty. Eight aspects of the neurological deficit present just before randomization were recorded. The computer system in the randomization centre employed a validated algorithm to assign the patient to one of the four infarct syndromes: lacunar syndrome (LACS), partial anterior circulation syndrome (PACS), total anterior circulation syndrome (TACS) and posterior circulation syndrome (POCS). We assessed the localization and extent of the recent infarction on available CT scans and correlated these with the computer-assigned OCSP category. CT scans were available for 558/759 (74%) of the patients randomized in Poland. In 458 (82%) of cases, CT was carried out in the first 24 h after the onset of stroke. In 444 (80%) scans, a recent infarct was visible. These radiological lesions were appropriate to the clinical classification in 56% of patients with TAGS, 73% with PALS, 61% with LACS and 59% with POCS. In Polish centres in IST, amongst the patients with infarction visible on CT, the OCSP subtype predicted the size and site of the infarct in about two-thirds of cases. These data suggest that, provided its limitations are taken into account, the classification can be usefully applied in multicentre clinical trials (or epidemiological studies) and to aspects of the routine clinical care of patients with acute stroke.